Train AI to Earn, an article to understand the AI data crowdsourcing network PublicAI
Invested by Saudi Telecom Group, Solana Foundation, NEAR Foundation, YGG, and others.Author: momo, ChainCatcher
Recently, the decentralized AI data training platform PublicAI completed an $8 million Series A financing round. Notably, the investors include well-known institutions and project parties such as Saudi Telecom Group, YGG, NEAR Foundation, and Whitewater Labs.
In January 2024, PublicAI also secured $2 million in funding led by IOBC Capital, with participation from Solana Foundation, Foresight Ventures, and Everstate Capital.
What kind of AI + Crypto product does PublicAI aim to create? What progress and participation opportunities are there? This article provides a brief overview.
Train t o Earn , AI Data Crowdsourcing Network in the Web3 Era
The traditional internet era has given rise to well-known data crowdsourcing platforms like Amazon Mechanical Turk and Scale AI.
These data crowdsourcing platforms connect two core parties: one is the data demand side, such as AI companies and research institutions that need a large amount of data to be labeled, cleaned, or classified; the other is the data contribution side, where the crowdsourcing platform breaks down the data tasks that need manual processing into smaller parts, distributes them to "part-timers" on the platform, and pays corresponding rewards.
PublicAI reworks these traditional data crowdsourcing models using blockchain technology and Web3 economic incentive models, focusing on the AI field.
On one hand, PublicAI provides high-quality, customized data labeling and processing services for AI companies that require a large amount of data for model training. On the other hand, PublicAI has established a decentralized network that allows users to train AI models and earn rewards by completing data tasks through the Train AI To Earn incentive model.
PublicAI currently operates through a three-layer architecture.
DataHub: This is the core crowdsourcing engine where contributors and validators collaborate. Uploaders submit data, and voters assess data quality through a consensus model.
Data Hunters: AI node operators initiate on-chain AI activities at this level, assisting in content validation and generation.
Blockchain and Smart Contract Layer: The blockchain layer ensures fair reward distribution and data validation through smart contracts, automating the platform.
Latest Progress and Participation Opportunities of PublicAI
PublicAI is currently in the testnet phase. Users can earn $PUBLIC airdrop points by completing social tasks in DataHub, becoming data uploaders, voters, etc.
PublicAI will release some upload tasks, and users with more than 100 fans can participate, generally required to submit compliant text or audio data.
Participants can also become voters, review uploaded data, and vote to determine whether it meets the activity standards, participating in consensus validation (Byzantine Fault Tolerance). They earn $PUBLIC airdrop points based on validation accuracy and consensus consistency, with higher accuracy receiving more rewards.
Additionally, participants can upload web data and complete related tasks to earn points through PublicAI's DataHunter Google plugin.
According to official data from PublicAI, the platform has attracted over 900,000 "workers" and 300,000 validators, with datasets exceeding 100T.
Background of the PublicAI Team
The core team members of PublicAI have backgrounds in traditional Web2 large companies.
PublicAI co-founder and Chief Operating Officer Kenji Narushima graduated from Tokyo University of Foreign Studies and obtained an MBA from Tsinghua University. Kenji Narushima has over 18 years of work experience, starting his career in business development and investment at Mitsubishi Corporation in Japan, and has also worked at well-known companies and institutions such as AWS and Forbes. He has served as Chief Strategy Officer and Chief of Staff at Carro, the largest used car market in Southeast Asia, and held a senior director position at AI company Ridecell.
PublicAI co-founder and CTO Steven Wong holds a PhD in machine learning from the Chinese University of Hong Kong and has been a lecturer for blockchain courses at the same university. He has also worked as a senior software engineer at NetEase Games.
PublicAI's Chief Marketing Officer Jordan Gray has previously worked at Google and the Near Foundation.
